Project 1 There is a need to identify and develop more favorable therapeutic index-based cancer preventive interventions. The overall goal of this project is to test the hypothesis that substituting iu3 fatty acids for 106 fatty acids in colorectal mucosal membranes will sufficiently modify the ratio of eicosapentaenoic acid (EPA) to arachidonic acid (AA) available to prostaglandin-H synthases (PGHS)-1 and 2 to reduce local prostaglandin (PG)E2 concentrations. Reduction of colonic mucosal PGE2 will reduce the carcinogenesis stress and ultimately reduce the development of neoplasia in the colonic epithelium. We also hypothesize that the dose of fish oil sufficient to reduce local PGE2 can be predicted by plasma EPAiAA ratio and thus be individualized. The reduction of PGE2 concentrations sufficient to reduce proliferation and enhance apoptosis of the colorectal mucosal crypt may be searched. Aim #1 will define tlie dose response to dietary fish oil in mice and the relationshipo between the plasma and colonic mucosal EPA:AA ratio upon reduction of PGE2 in colonic mucosa of PGHS-1 and -2 wild type mice will be fed diets formulated to match multiple human fatty acid intakes. The plasma and colonic mucosal EPA:AA ratios and PGE2 concentrations in colonic mucosa and urine will be assayed. Data in Aim 1 are necessary to define the design of a Phase I clinical trial proposed in Aim 2. The clincal trial will determine if fish oil supplementation in humans on a normal diet can produce a colorectal mucosal and plasma EPAiAA ratio defined in the mouse models that will reduce colorectal mucosal PGE2, reduce colorectal crypt proliferation index and enhance apoptosis indicies. A Bayesian driven biomarker adaptive phase I design individualizes dose on the basis of individual biomarker response. The data obtained in this project will determine the feasibility, future design and biomarker endpoints of Phase II clinical trials of u)3 fatty acids as potential preventives of colorectal adenocarcinoma.